Search Results (160)

Industrial Vision Inspection Systems (VISs) often struggle to adapt to increasing variability of modern manufacturing due to the inherent rigidity of their hardware architectures. Although the Reconfigurable Manufacturing System (RMS) paradigm was introduced in the early 2000s to overcome these limitations, designing such reconfigurable machines remains a complex, expert-dependent, and time-consuming task. This is primarily due to the lack of structured methodologies and the reliance on trial-and-error processes. In this context, this study proposes a novel theoretical framework to facilitate the design of fully reconfigurable handling systems for VISs, with a particular focus on fixture design. The framework is grounded in Model-Based Definition (MBD), embedding semantic information directly into the 3D CAD models of the inspected product. As an additional contribution, a general hardware architecture for the inspection of axisymmetric components is presented. This architecture integrates an anthropomorphic robotic arm, Numerically Controlled (NC) modules, and adaptable software and hardware components to enable automated, software-driven reconfiguration. The proposed framework and architecture were applied in an industrial case study conducted in collaboration with a leading automotive half-shaft manufacturer. The resulting system, implemented across seven automated cells, successfully inspected over 200 part types from 12 part families and detected more than 60 defect types, with a cycle below 30 s per part. Full article

As a critical precision transmission element in numerical control (NC) machines, the linear rolling guide (LRG) suffers from surface wear degradation, which significantly impairs machining accuracy and operational reliability. Despite its importance, effective identification methods for LRG degradation remain limited. In this study, a hybrid approach combining multi-scale fuzzy entropy (MFE) with a gray wolf-optimized random forest (GWO-RF) algorithm was proposed to identify the surface wear state of the LRG. Preload degradation and vibration signals were collected at three surface wear stages throughout the LGR’s service life. The vibration signals were decomposed and reconstructed using complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), followed by multi-scale fuzzy entropy analysis of the reconstructed signals. After dimensionality reduction via kernel principal component analysis (KPCA), the processed features were fed into the GWO-RF model for classification. Experimental results demonstrated a recognition accuracy of 97.9%. Full article

To address the growing demand for environmentally friendly flexible sensors, here, a composite hydrogel of nanocellulose (NC) and polyvinyl alcohol (PVA) was designed and fabricated using Camellia oleifera shells as a sustainable alternative to petroleum-based raw materials. Firstly, NC was extracted from Camellia oleifera shells and modified with 2-chloropropyl chloride to obtain a nanocellulose-based initiator (Init-NC) for atomic transfer radical polymerization (ATRP). Subsequently, sulfonyl betaine methacrylate (SBMA) was polymerized by Init-NC initiating to yield zwitterion-functionalized nanocellulose (NC-PSBMA). Finally, the NC-PSBMA/PVA hydrogel was fabricated by blending NC-PSBMA with PVA. A Fourier transform infrared spectrometer (FT-IR), proton nuclear magnetic resonance spectrometer (¹H-NMR), X-ray diffraction (XRD), scanning electron microscope (SEM), transmission electron microscope (TEM), universal mechanical testing machine, and digital source-meter were used to characterize the chemical structure, surface microstructure, and sensing performance. The results indicated that: (1) FT-IR and ¹H NMR confirmed the successful synthesis of NC-PSBMA; (2) SEM, TEM, and alternating current (AC) impedance spectroscopy verified that the NC-PSBMA/PVA hydrogel exhibits a uniform porous structure (pore diameter was 1.1737 μm), resulting in significantly better porosity (15.75%) and ionic conductivity (2.652 S·m⁻¹) compared to the pure PVA hydrogel; and (3) mechanical testing combined with source meter testing showed that the tensile strength of the composite hydrogel increased by 6.4 times compared to the pure PVA hydrogel; meanwhile, it showed a high sensitivity (GF = 1.40, strain range 0–5%; GF = 1.67, strain range 5–20%) and rapid response time (<0.05 s). This study presents a novel approach to developing bio-based, flexible sensing materials. Full article

In computer numeric controller (CNC) milling machining, air-cutting, where the tool moves without engaging the material, will reduce the machining efficiency. This study proposes a novel methodology to detect and minimize non-productive (air-cutting) time in real-time using spindle load monitoring, vibration signal analysis, and NC code tracking. A logic-based benders decomposition (LBBD) approach was used to optimize toolpath segments by analyzing air-cutting occurrences and dynamically modifying the NC code. Two optimization strategies were proposed: increasing the feedrate during short air-cutting segments and decomposing longer segments using G00 and G01 codes with positioning error compensation. A human–machine interface (HMI) developed in C# enables real-time monitoring, detection, and minimization of air-cutting. Experimental results demonstrate up to 73% reduction of air-cutting time and up to 42% savings in total machining time, validated across multiple scenarios with varying cutting parameters. The proposed methodology offers a practical and effective solution to enhance CNC milling productivity. Full article

Cataract remains the leading cause of blindness worldwide, and the number of individuals affected by this condition is expected to rise significantly due to global population ageing. Early diagnosis is crucial, as delayed treatment may result in irreversible vision loss. This study explores and presents a mobile application for Android devices designed for the detection of cataracts using deep learning models. The proposed solution utilizes a multi-stage classification approach to analyze ocular images acquired with a slit lamp, sourced from the Nuclear Cataract Database for Biomedical and Machine Learning Applications. The process involves identifying pathological features and assessing the severity of the detected condition, enabling comprehensive characterization of the NC (nuclear cataract) of cataract progression based on the LOCS III scale classification. The evaluation included a range of convolutional neural network architectures, from larger models like VGG16 and ResNet50, to lighter alternatives such as VGG11, ResNet18, MobileNetV2, and EfficientNet-B0. All models demonstrated comparable performance, with classification accuracies exceeding 91–94.5%. The trained models were optimized for mobile deployment, enabling real-time analysis of eye images captured with the device camera or selected from local storage. The presented mobile application, trained and validated on authentic clinician-labeled pictures, represents a significant advancement over existing mobile tools. The preliminary evaluations demonstrated a high accuracy in cataract detection and severity grading. These results confirm the approach is feasible and will serve as the foundation for ongoing development and extensions. Full article

Alzheimer’s disease (AD) is a leading cause of dementia worldwide. As current diagnostic approaches remain limited in sensitivity and accessibility, there is a critical need for novel, non-invasive biomarkers aiding early detection. Non-coding RNAs (ncRNAs), including long non-coding RNAs (lncRNAs), PIWI-interacting RNAs (piRNAs), and small nucleolar RNAs (snoRNAs), have emerged as promising candidates due to their regulatory roles in gene expression and association with diseases. In this study, we systematically profiled ncRNA expression from RNA sequencing data of 48 AD and 22 control blood tissue samples, aiming to evaluate their utility as biomarkers for AD classification. Differential expression analysis revealed widespread dysregulation of lncRNAs and piRNAs, with over 5000 lncRNAs and nearly 1000 piRNAs significantly upregulated in AD. Weighted gene co-expression network analysis (WGCNA) identified multiple ncRNA modules associated with the AD phenotype. Using supervised machine learning approaches, we evaluated the diagnostic potential of ncRNA expression profiles, including single-gene, multi-gene, and module-level models. Random Forest models trained on individual genes identified 121 ncRNAs with AUROC > 0.8. Feature importance analysis emphasized ncRNAs such as lnc-MYEF2-3, lnc-PRKACB2, and HBII-115 as major contributors to diagnostic accuracy. These findings support the potential of ncRNA signatures as reliable and non-invasive biomarkers for AD diagnosis. Full article

Federated learning is a distributed machine learning approach where end users train local models with their own data and combine model updates on a reliable server to create a global model. Despite its advantages, this distributed structure is vulnerable to attacks as end users keep their data and training process private. Current defense mechanisms often fail when facing different attack types or high percentages of malicious participants. This paper proposes a new defense algorithm called Neuron-Centric Federated Learning Defense (NC-FLD), a novel approach that dynamically identifies and analyzes the most significant neurons across model layers rather than examining entire gradient spaces. Unlike existing methods that analyze all parameters equally, NC-FLD creates feature vectors from specifically selected neurons that show the highest training impact, then applies dimensionality reduction to enhance their discriminative features. We conduct experiments with various attack scenarios and different malicious participant rates across multiple datasets (CIFAR-10, F-MNIST, and MNIST). Additionally, we perform simulations on the GTSR dataset as a real-world application. Experimental results demonstrate that NC-FLD successfully defends against diverse attack scenarios in both IID and non-IID dataset distributions, maintaining accuracy above 70% with 40% malicious participation, a 5–15% improvement over the state-of-the-art method, showing enhanced robustness across diverse data distributions while effectively mitigating the impacts of both data and model poisoning attacks. Full article

This study proposes an asynchronous brain–computer interface (BCI) framework based on steady-state motion visual evoked potentials (SSMVEPs), designed to enhance the accuracy and robustness of control state recognition. The method integrates filter bank common spatial patterns (FBCSPs) and filter bank canonical correlation analysis (FBCCA) to extract complementary spatial and frequency domain features from EEG signals. These multimodal features are then fused and input into a dual-classifier structure consisting of a support vector machine (SVM) and extreme gradient boosting (XGBoost). A weighted fusion strategy is applied to combine the probabilistic outputs of both classifiers, allowing the system to leverage their respective strengths. Experimental results demonstrate that the fused FB(CSP + CCA)-(SVM + XGBoost) model achieves superior performance in distinguishing intentional control (IC) and non-control (NC) states compared to models using a single feature type or classifier. Furthermore, the visualization of feature distributions using UMAP shows improved inter-class separability when combining FBCSP and FBCCA features. These findings confirm the effectiveness of both feature-level and classifier-level fusion in asynchronous BCI systems. The proposed approach offers a promising and practical solution for developing more reliable and user-adaptive BCI applications, particularly in real-world environments requiring flexible control without external cues. Full article

Hardox 500 is a special low-alloy, martensitic steel possessing extraordinary wear resistance, high hardness, and high ductility; thus, it has been widely used in many industrial applications. Nevertheless, this type of steel has a low machinability and is grouped among the difficult-to-machine materials. Hence, this paper’s objective was to study its hard milling performance under minimum quantity cooling lubrication (MQCL) conditions using an Al₂O₃/MoS₂ hybrid nano cutting oil. The Box–Behnken response surface methodology was used to investigate the effects of the nanoparticle concentration (NC), cutting speed (v), and feed rate (f) on the total cutting force F and cutting force coefficient F_y/F_z. The obtained results indicate that the cutting efficiency of Hardox 500 steel was improved thanks to the enhancement in cooling lubrication from the MQCL using the Al₂O₃/MoS₂ hybrid nano cutting oil. The applicability of vegetable oil and coated carbide inserts is thus extended to the hard milling of difficult-to-cut materials. Moreover, the provision of the appropriate ranges and optimal set of investigated variables obtained in this paper will be useful guides for technologists and further studies. Concretely, NC = 0.5–0.7%, v = 110–115 m/min, and f = 0.08–0.10 mm/tooth are the optimal set for the total cutting force F, while NC = 0.5%, v = 138–140 m/min, and f = 0.08–0.09 mm/tooth are suggested for the cutting force coefficient F_y/F_z. Full article

With the development of intelligent manufacturing technology, the manufacturing industry is gradually realizing intelligent production. Especially for metal cutting with extremely complex processes, it is of great significance to realize intelligence. Taking the cutting process of aero-engine typical low-rigidity parts as the main line, this article builds an intelligent processing architecture based on a big data platform, which includes customized design of cutting tools, intelligent optimization of cutting parameters, simulation of cutting conditions, and online monitoring and control of cutting processes. At the same time, the realization of related key technologies is explained. Then, this article introduces in detail the intelligent decision-making process based on deep learning, the customized tool design process based on structural features, the simulation process of cutting based on geometric features of parts, as well as the monitoring and control process of Numerical Control (NC) machining based on condition perception. In addition, based on the processing requirements and difficulties of specific parts, formulate a specific intelligent implementation plan under this processing mode. Through the implementation of the above architecture and key technologies, the cutting processing system can automatically optimize the cutting parameters according to real-time working conditions and adjust its own cutting conditions. At the same time, machine tool condition, cutting tool condition, and low-rigidity part condition are real-time monitored to achieve high-precision, efficient, intelligent, and precise cutting of low-rigidity parts. The proposed architecture can provide a reference model for the research and application of intelligent cutting technology for low-rigidity parts. Full article

The monolithic composite action of structures relies on the interface bond strength between concrete and the repair material. This study uses explainable deep learning techniques to evaluate the ultimate strength capacity (Us) of U-shaped normal concrete (NC) strengthened with polyurethane grouting (PUG) materials. Machine learning algorithms (ML) such as Long Short-Term Memory (LSTM), Random Forest (RF), and Wide Neural Network (WNN) models were developed to estimate Us by considering five input parameters: the initial crack strength (Cs), thickness of the grouting materials (T), mid-span deflection (λ), and peak applied load (P). The results indicated that LSTM models, particularly LSTM-M2 and LSTM-M3, demonstrated superior predictive accuracy and consistency in both the calibration and verification phases, as evidenced by high Pearson’s correlation coefficients (PCC = 0.9156 for LSTM-M2) and Willmott indices (WI = 0.7713 for LSTM-M2), and low error metrics (MSE = 0.0017, RMSE = 0.0418). The SHAP (SHapley Additive exPlanations) analysis showed that the thickness of the grouting materials and maximum load were the most significant parameters affecting the ultimate capacity of the composite U-shaped specimen. The RF model showed moderate improvements, with RF-M3 performing better than RF-M1 and RF-M2. The WNN models displayed varied performance, with WNN-M2 performing poorly due to significant scatter and deviation. The findings highlight the potential of LSTM models for the accurate and reliable prediction of the ultimate strength of composite U-shaped specimens. Full article

The rapid advancement of sustainable materials has driven the need for high-performance polymer nanocomposites with superior mechanical, thermal, and structural properties. In this study, a novel RPET/PA-11/Joncryl^® nanocomposite reinforced with halloysite nanotubes (HNTs) is developed for the first time, marking a significant breakthrough in polymer engineering. Six different proportions of HNT (0, 1, 2, 3, 4, and 5 phr) are introduced to the blend of rPET/PA-11/Joncryl^® through a twin-screw extruder and injection moulding machine. The incorporation of HNTs into the RPET/PA-11 matrix, coupled with Joncryl^® as a compatibilizer, results in a synergistic enhancement of material properties through improved interfacial adhesion, load transfer efficiency, and nanoscale reinforcement. Comprehensive characterization reveals that the optimal formulation with 2 phr HNT (NCS-H2) achieves remarkable improvements in tensile strength (56.14 MPa), flexural strength (68.34 MPa), and Young’s modulus (895 MPa), far exceeding conventional polymer blends. Impact resistance reaches 243.46 J/m, demonstrating exceptional energy absorption and fracture toughness. Thermal analysis confirms enhanced stability, with an onset degradation temperature of 370 °C, attributing the improvement to effective matrix–filler interactions and restricted chain mobility. Morphological analysis through FESEM validates uniform HNT dispersion at optimal loading, eliminating agglomeration-induced stress concentrators and reinforcing the polymer network. The pioneering integration of HNT into RPET/PA-11/Joncryl^® nanocomposites not only bridges a critical gap in sustainable polymers but also establishes a new benchmark for polymer nanocomposites. This work presents an eco-friendly solution for engineering applications, offering mechanical robustness, thermal stability, and recyclability. The results form the basis for next-generation high-performance materials for industrial use in automotive, aerospace, and high-strength structural applications. Full article

Sarcopenia, a condition characterised by the progressive decline in skeletal muscle mass and function, presents significant challenges in geriatric healthcare. Despite advances in its management, complex etiopathogenesis and the heterogeneity of diagnostic criteria underlie the limited precision of existing assessment methods. Therefore, efforts are needed to improve the knowledge and pave the way for more effective management and a more precise diagnosis. To this purpose, emerging technologies such as artificial intelligence (AI) can facilitate the identification of novel and accurate biomarkers by modelling complex data resulting from high-throughput technologies, fostering the setting up of a more precise approach. Based on such considerations, this review explores AI’s transformative potential, illustrating studies that integrate AI, especially machine learning and deep learning, with heterogeneous data such as clinical, anthropometric and molecular data. Overall, the present review will highlight the relevance of large-scale, standardised studies to validate biomarker signatures using AI-driven approaches. Full article

Objectives: Trauma to maxillary incisors is frequent, and requires timely, conservative management for optimal prognosis. This in vitro study evaluated the fracture resistance (FR) and orthodontic bracket bond strength (BS) of incisors following incisal fragment reattachment using various restorative techniques. Materials and Methods: Two independent tests—FR testing (Newtons) and BS testing (megapascals)—were conducted. Eighty intact human maxillary central incisors (n = 40/test), standardized in size and shape using a digital caliper (Mitutoyo, ±0.01 mm), were embedded in acrylic resin and numbered. An uncomplicated crown fracture was induced in 64 teeth (n = 32/test), and the teeth were randomly assigned (simple randomization using Excel’s RAND function) to five groups (n = 8/group/test): (1) intact teeth (negative control, NC); (2) nanohybrid composite buildup using Filtek Z250 and Single Bond 2 (positive control, CB); (3) fragment reattachment using flowable composite (Filtek Supreme, FL); (4) reattachment with a palatal veneer using a nanohybrid composite (PV); and (5) reattachment reinforced with a polyethylene fiber band (Ribbond Inc., RB). In BS testing groups, stainless steel orthodontic brackets (PINNACLE) were bonded using Transbond XT, centered over the fracture line. Light curing was performed using an LED unit (Mini LED Standard, Acteon, 1250 mW/cm², 20 s/bond, 40 s/composite, 2 mm curing tip distance). Specimens were stored in distilled water at room temperature for 24 h before reattachment. FR and BS were evaluated using a universal testing machine (Instron) until failure. Failure modes were analyzed, and data were statistically evaluated using one-way ANOVA, Tukey’s post hoc test, and Pearson’s correlation analysis. Results: Significant differences were observed among groups for both FR and BS (p < 0.05). The NC group exhibited the highest FR (514.4 N) and BS (17.6 MPa). The RB group recorded the second-highest FR (324.6), followed by the PV (234.6), CB (224.9), and FL (203.7) groups. The CB group demonstrated the second-best BS (16.6), followed by the RB (15.2), FL (13.4), and PV (6.5) groups. FR and BS were negatively correlated. Mixed failures predominated in the reattachment groups, except for the PV group, which showed mainly adhesive failures. In BS testing, mixed failures dominated in the NC and CB groups, while adhesive failures predominated in the PV and FL groups. Conclusions: Ribbond reinforcement improves the mechanical performance of reattached incisal fragments, and composite buildup may provide more reliable bracket bonding than fragment reattachment. Clinical Relevance: In cases where biomimetic, minimally invasive reattachment is indicated, Ribbond fiber reinforcement appears to offer a reliable restorative solution. Full article

The strong correlation between evoked potentials (EPs) and American Spinal Injury Association (ASIA) scores in individuals with spinal cord injury (SCI) suggests that EPs may serve as reliable predictive markers for rehabilitation progress. Numerous studies have confirmed a relationship between variations in somatosensory evoked potentials (SSEPs) and ASIA scores, especially in the early stages of SCI. Machine learning’s (ML’s) increasing importance in medicine is driven by the growing availability of health data and improved algorithms. It enables the creation of predictive models for disease diagnosis, progression prediction, personalized treatment, and improved healthcare efficiency. Data-driven approaches can significantly improve patient care, reduce costs, and facilitate personalized medicine. The meticulous analysis of medical data is crucial for timely disease identification, leading to effective symptom management and appropriate treatment. This study applies artificial intelligence to identify predictors of SCI progression, as measured by the disability index, ASIA impairment scale (AIS), and final motor recovery. We aim to clarify the prognostic role of electrophysiological testing (SSEPs, MEPs, and nerve conduction studies (NCSs)) in SCI. We analyzed data from a medical database of 123 records. We developed an ML-based intelligent system, utilizing ensemble algorithms combining decision trees and neural network approaches, to predict SCI recovery. Our evaluation showed SEP accuracies of 90% for motor recovery prediction and 80% for AIS scale determination, comparable to full electrophysiology evaluation accuracies of 93% and 89%, respectively, and generally superior results compared to MEP and NCS results. EPs emerged as the best predictors, comparable to a comprehensive electrophysiology assessment, significantly improving accuracy compared to clinical findings alone. An electrophysiological assessment, when available, increased overall accuracy for final motor recovery prediction to 93% (from a maximum of 75%) and, for ASIA score determination, to 89% (from a maximum of 66%). Further validation is needed with a larger dataset. Future research should validate that sensory electrophysiology assessment is a less expensive, portable, and simpler alternative to other prognostic tests and more effective than clinical assessments, like the AIS, biomarker for SCI, and personalized rehabilitation planning. Full article